ABSTRACT
This chapter introduces methods for answering the critical question that arises in planning an experiment, namely, how large must the sample sizes be to make it likely that an effect of a given projected size will be detected. It proceeds from the simplest case of this general linear model to more and more complex forms. The chapter considers a one-group situation, a two-group situation, and then situations involving three or more groups of subjects. It assumes that the samples represent independent groups of participants and that these groups differ along a single dimension or factor. The model comparison approach the authors emphasize in the chapter views Analysis of variance (ANOVA) from the more general perspective of being a method that researchers can use in deciding what linear model is appropriate for describing the data obtained in a study.
